The below post was written by Rob Luker. Rob used to write for Blueshirt Banter, but retired for a while. He sent me this and I was more than happy to post it. You should be following him on Twitter here. He is a great follow with tons of good hockey insight.
When Larry Brooks pointed out that the Rangers are expected to try to trade Nick Holden in his Tuesday, July 25th column, this got me texting and chatting with friends. Back on May 9th, fresh off the series loss to Ottawa, I tweeted the following:
Here's to hoping Gorton can somehow "moneyball" AV aka clean out his bad options. 5/8/18/15 would be addition by subtraction. #SerentiyNow
— Rob Luker (@RLuker12) May 10, 2017
So far, I believe Jeff Gorton has done a pretty solid job of doing just that to Alain Vigneault. While there are still risks of how AV could make poor personnel decisions, the roster is better constructed to limit those risks. If Holden is moved for any asset (even beyond a bottom six center), that takes away the potential for AV to use Holden in a larger-minute role, while subsequently opening up a slot for one of the prospects.
That all being said, many voices have covered both the overall offseason feelings and Vigneault’s potential deployment options, so what has prompted me to write for the first time in a long while? It was a comment from my St. Louis friend Kevin Lorenz. While discussing the team as a whole if Holden was moved, he said “it just seems like they all [NYR forwards] need to have a good year at the same time.”
Now, having lived in St. Louis from 2013 to 2016, I appreciate Kevin’s opinion on the Rangers because the Blues have been in an opposite situation than NYR has since 2013. Under AV, the Rangers have been a middle ten of the NHL team in terms of Corsi events (rates for and against, combined, meaning the Rangers played pretty fast paced hockey. They pushed the puck offensively, but were lapse defensively as well), but they were top ten in terms of scoring chance events. The percentages (CF%, SCF%, and xGF% – all at evens and regular season only) have swayed year to year depending on the roster, but the trading of scoring chances has been consistent.
The Blues, however, played prototypical Ken Hitchock hockey during this time. Bottom ten in the league when it came to Corsi and Scoring Chance events/60 (basically playing dull hockey with minimal offense either way) all while controlling the percentages mentioned above consistently. Finally, when it comes to goaltending, let’s just say that Blues fans are still praying that their own Lundqvist comes along eventually.
Usually when the term regression is brought out the infamous PDO acronym follows not too long after. And while shot and save percentages will always give us outlier examples every season, advances in other areas have given us a little more insight than just basic percentages. Right now (not that this is a revelation of any sort) I am a big fan of using expected goals when looking at individual player performances at a season level.
Specifically, I tend to start with “Surplus Goals,” which is nicely visualized on Cole Anderson’s Player Compare App. Surplus Goals simply equals actual individual Goals/60 (G/60) minus expected individual Goals/60 (ixG/60), all at even strength (note: the data does include playoff games for each year/player). Essentially, if a player is wildly positive or negative, it would be prudent to expect that the same performance is likely not to happen in 2017-18. Here’s how the whole league looks when plotted:
Some necessary stats on how this works (if stats aren’t your thing, you can skip this paragraph and go to the explanation below): For our sample size of 841 players in 2016-17 (x axis), the average is 0.23 surplus goals (y axis), while the median is -0.1 and 2.58 equals one standard deviation. 77.5% (652 players) fall within one standard deviation while 93.1% (783 players) fall within two deviations (-5.16 to 5.16). The top three at each end looks like the following (hint: fantasy hockey tip):
Basically, read this just like you would PDO: If it’s higher that means the player scored much more per 60 minutes than the expected goals model had him scoring, and vice versa. Naturally, any speculation needs to come with a healthy amount of context (Laine probably will continue to shoot pretty damn well for a while), but it’s nonetheless interesting. How does the updated NYR roster look?
Finally, the train of thought that started with moving Nick Holden can result in some analysis. Unsurprisingly, Michael Grabner is at the top of NYR’s Surplus Goals list after his 27 goal season (26 at even strength) and is also two standard deviations away from zero at 7.6. That is further evidence that he will likely not repeat last season’s success, as many have already stated.
Another expected name in Nick Holden follows Grabner, but then we get into some interesting names who most all think had good seasons. Chris Kreider, J.T. Miller, and Mika Zibanejad all fall outside of one standard deviation, meaning they are three of 115 NHL players on the positive side of Surplus Goals higher than 2.58. Basically that means that we have some reason to believe that they might regress next season. It’s not a guarantee, but there’s some evidence behind it.
Let’s look at each players individual expected and actual goal rates over their careers to see if they’re looking at a regression risk (and what that means for the NYR forward group). Note: anything near or above 1.0 ixG/60 or G/60 (the blue and red lines) is very much top end for the NHL.
Kreider has generally been on the rise since his 23 game rookie season in 2012-13. This past season, he saw a bump up in ice time while recording career highs in shooting percentage, total shot attempts (corsi), and shots on goal. Going into his age 26 season, it will be interesting to see if Kreider can continue the trend and crack 30 goals, or if he sees a slight bump back given that his expected goal rates have been slightly less than his actual goal rates.
Miller’s actual goal rate has been also on the rise since his rookie season while his individual expected rate has hovered around 0.5 consistently. His counting stats have gone up as his minutes have risen (career high in 2016-17), so the same question is essentially posed for him as it is Kreider when it comes to 2017-18.
Zibanejad has consistently outscored his expected rates since he started playing real forward minutes in the NHL. He masked a low-volume shooting year in 2014-15 with a higher shooting percentage, and it again happened to an extent last season. The defacto number one NYR center will be another one to watch.
For comparisons sake, here are Crosby and Ovechkin:
The theme here is that three important NYR forwards have consistently outscored their expected rates the past few seasons. From a statistics standpoint, this very much worries me as all players will have “down” years happen to them (hence Kevin’s original quote).
However that doesn’t mean they will all drop next year, as it could certainly come when they decline past 26/27, but it’s still possible. On the positive side, the coaching staff has presumably done well to have NYR consistently provide offense since 2013, as I mentioned above (given the scoring chance rates).
For Kreider, Miller, and Zibanejad, the outscoring of their expected rates (Kevin Hayes and Mats Zuccarello have also outscored their rates in all of their seasons) is most likely a combination of surrounding personnel and coaching system – a credit to Sather, Gorton, and the organization as a whole. This at least provides some (unquantifiable) relief from a statistical standpoint of regression with these forwards. Remember: Talent and systems matter in all this, and it’s almost impossible to quantify. This is an exercise in using what we have and attempting to predict what may happen next year.
As I alluded to above, I think the roster the Rangers have today is better off than the one they finished with on May 9th. Another natural center is preferred, but if Miller transitions and another winger steps up, in theory a healthy Rangers squad would have a top ten-ish offense, a top ten-ish defense, and Lundqvist. This is all contingent on coaching staff usage and injuries, plus the annual PDO roulette spin that is every NHL season.
What could possibly go wrong?"Guest Post: NY Rangers Forwards, A Mix of Skill and System",